[llvm] [VPlan] Introduce ComputeReductionResult VPInstruction opcode. (PR #70253)

[Florian Hahn via llvm-commits]

================
@@ -9130,55 +9009,79 @@ void LoopVectorizationPlanner::adjustRecipesForReductions(
     for (VPRecipeBase &R :
          Plan->getVectorLoopRegion()->getEntryBasicBlock()->phis()) {
     VPReductionPHIRecipe *PhiR = dyn_cast<VPReductionPHIRecipe>(&R);
-    if (!PhiR || PhiR->isInLoop())
+    if (!PhiR)
       continue;
 
     const RecurrenceDescriptor &RdxDesc = PhiR->getRecurrenceDescriptor();
-    auto *Result = PhiR->getBackedgeValue()->getDefiningRecipe();
     // If tail is folded by masking, introduce selects between the phi
     // and the live-out instruction of each reduction, at the beginning of the
     // dedicated latch block.
-    if (CM.foldTailByMasking()) {
+    auto *OrigExitValue = PhiR->getBackedgeValue();
+    auto *ExitValue = OrigExitValue->getDefiningRecipe();
----------------
fhahn wrote:

Updated, thanks!

https://github.com/llvm/llvm-project/pull/70253